Fault identification method and apparatus for transformer, device, and medium

Abstract

Disclosed in embodiments of the present application are a fault identification method and apparatus for a transformer, a device, and a medium. The method comprises: acquiring concentration data of dissolved gas in transformer oil; and processing the concentration data on the basis of an optimized probabilistic neural network model to obtain a fault type of a transformer, wherein the optimized probabilistic neural network model is determined on the basis of a target smooth factor parameter, the target smooth factor parameter is determined from among candidate smooth factor parameters by means of an artificial jellyfish search algorithm, and the candidate smooth factor parameters are smooth factor parameters meeting a fitness condition.

Description

Transformer fault identification method, device, equipment and medium

[0001] This application claims priority to the Chinese patent application filed with the China Patent Office on July 31, 2023, with application number 202310957273.4, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present application relates to the technical field of transformers, and for example, to a method, device, equipment, and medium for identifying transformer faults. Background Art

[0003] Dissolved gas analysis (DGA) in transformer oil is an important method for diagnosing internal transformer faults. Currently, one approach in the industry uses a modified three-ratio method to identify transformer faults; another approach combines artificial intelligence algorithms with DGA technology to diagnose transformer faults. However, current AI-based fault identification methods suffer from poor accuracy.

[0004] Summary of the Invention

[0005] The present application provides a transformer fault identification method, apparatus, equipment and medium, which can accurately identify the fault type of the transformer.

[0006] According to one aspect of the present application, a transformer fault identification method is provided, the method comprising:

[0007] Obtain the concentration data of dissolved gas in transformer oil;

[0008] The concentration data is processed based on an optimized probabilistic neural network model to obtain the fault type of the transformer; wherein the optimized probabilistic neural network model is determined according to a target smoothing factor parameter; the target smoothing factor parameter is determined from candidate smoothing factor parameters by an artificial jellyfish search algorithm; the candidate smoothing factor parameter is a smoothing factor parameter that meets a fitness condition.

[0009] According to another aspect of the present application, a transformer fault identification device is provided, comprising:

[0010] A concentration data acquisition module is configured to acquire concentration data of dissolved gas in transformer oil;

[0011] A fault type determination module is configured to process the concentration data based on an optimized probabilistic neural network model to obtain the fault type of the transformer; wherein the optimized probabilistic neural network model is determined according to a target smoothing factor parameter; the target smoothing factor parameter is determined from candidate smoothing factor parameters by an artificial jellyfish search algorithm; and the candidate smoothing factor parameter is a smoothing factor parameter that meets a fitness condition.

[0012] According to another aspect of the present application, an electronic device is provided, comprising:

[0013] at least one processor; and

[0014] a memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the transformer fault identification method described in any embodiment of the present application.

[0016] According to another aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the transformer fault identification method described in any embodiment of the present application when executed. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] FIG1 is a flow chart of a transformer fault identification method provided according to an embodiment of the present application;

[0018] FIG2 is a flow chart of another transformer fault identification method provided according to an embodiment of the present application;

[0019] FIG3 is a schematic structural diagram of a transformer fault identification device provided according to an embodiment of the present application;

[0020] FIG4 is a schematic structural diagram of an electronic device for implementing a transformer fault identification method according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] It should be noted that the terms "first", "second", "target", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] FIG1 is a flowchart of a transformer fault identification method provided in an embodiment of the present application. The embodiment of the present application is applicable to identifying internal faults of oil-immersed transformers. The method can be executed by a transformer fault identification device, which can be implemented in the form of hardware and / or software. The transformer fault identification device can be configured in an electronic device with data processing capabilities. As shown in FIG1 , the method includes:

[0023] S110, obtaining concentration data of dissolved gas in transformer oil.

[0024] Among them, the transformer can be an oil-immersed transformer. After various faults occur in this type of transformer, the concentration of dissolved gas in the transformer oil will also change. Therefore, the fault type of the transformer can be determined by the concentration data of dissolved gas in the transformer oil.

[0025] Dissolved gas in oil refers to the gas dissolved in the transformer oil in a molecular state. The sources of dissolved gas in oil are mainly air dissolution, gas generated by normal operation, and gas generated by internal faults. The main components of dissolved gas in transformer oil are N2, O2, H2, CH4, C2H2, C2H4, C2H6, C3H6, C3H8, CO, CO2, etc. For example, the concentration data of dissolved gas in transformer oil obtained in the embodiments of the present application can be the concentration data of H2, CH4, C2H6, C2H4, and C2H2.

[0026] For example, after detecting that the fault identification control has been triggered, an input box for the concentration of dissolved gas in oil can be displayed to obtain the concentration data of dissolved gas in transformer oil input by the staff through the input box. In another feasible solution, an input interface for the concentration of dissolved gas in oil is provided, and the input interface is connected to a concentration acquisition device for dissolved gas in oil, so that the concentration data of dissolved gas in oil can be read in real time. Of the above two solutions, one can obtain the concentration data of dissolved gas in transformer oil input by the staff. The concentration data can be relatively accurate concentration data after calibration, and the type of transformer fault can be accurately identified through subsequent steps; the other can realize the real-time acquisition of the concentration data of dissolved gas in transformer oil, and thus can detect in real time whether the transformer has a fault.

[0027] S120, processing the concentration data based on the optimized probabilistic neural network model to obtain the fault type of the transformer; the optimized probabilistic neural network model is determined according to a target smoothing factor parameter; the target smoothing factor parameter is determined from candidate smoothing factor parameters through an artificial jellyfish search algorithm; the candidate smoothing factor parameter is a smoothing factor parameter that meets a fitness condition.

[0028] The probabilistic neural network model is a type of machine learning model. Since this application identifies the corresponding transformer fault type based on dissolved gas concentration data in oil, it can be understood as a classification problem. Therefore, a probabilistic neural network model suitable for pattern classification can be used. After the probabilistic neural network model is established, it is optimized using the artificial jellyfish search algorithm to obtain the optimized probabilistic neural network model.

[0029] The smoothing factor parameter is an important parameter of the probabilistic neural network model. The optimization process of the probabilistic neural network model is the process of determining the smoothing factor parameter. In the embodiment of the present application, the smoothing factor parameter finally determined is called the target smoothing factor parameter, and the target smoothing factor parameter is determined from the candidate smoothing factor parameters.

[0030] For example, data on the concentration of dissolved gas in transformer oil is input into an optimized probabilistic neural network model. The optimized probabilistic neural network model processes the concentration data to generate an output result, which can be an output code. Different output codes correspond to different fault types. Exemplarily, the fault type includes at least one of the following: no fault, medium-low temperature overheating between 150°C and 700°C, high temperature overheating greater than 700°C, low-energy discharge, high-energy discharge, and partial discharge.

[0031] In an embodiment of the present application, a probabilistic neural network model is optimized based on an artificial jellyfish search algorithm to obtain an optimized probabilistic neural network model. For example, candidate smoothing factor parameters are selected based on the artificial jellyfish search algorithm to obtain target smoothing factor parameters, and then the optimized probabilistic neural network model is determined based on the target smoothing factor parameters. The candidate smoothing factor parameters are smoothing factor parameters that meet a fitness condition, i.e., the candidate smoothing factor parameters are smoothing factor parameters that have been screened, and this screening condition is the fitness condition.

[0032] Since the candidate smoothing factor parameters are smoothing factor parameters that meet the fitness conditions, that is, the candidate smoothing factor parameters are closer to the true value than other smoothing factor parameters, the optimal target smoothing factor parameters are determined from the candidate smoothing factor parameters, and then the optimized probabilistic neural network model is determined according to the target smoothing factor parameters, which makes the identification of fault types more accurate.

[0033] The technical solution of an embodiment of the present application includes: obtaining dissolved gas concentration data in transformer oil; processing the concentration data based on an optimized probabilistic neural network model to determine the fault type of the transformer; the optimized probabilistic neural network model is determined based on a target smoothing factor parameter; the target smoothing factor parameter is determined from candidate smoothing factor parameters using an artificial jellyfish search algorithm; the candidate smoothing factor parameters are smoothing factor parameters that meet fitness conditions. This technical solution determines the target smoothing factor parameter from the candidate smoothing factor parameters that meet the fitness conditions, and then determines the optimized probabilistic neural network model based on the target smoothing factor parameter, thereby improving the accuracy and efficiency of fault identification when using the model.

[0034] FIG2 is a flow chart of another transformer fault identification method provided in an embodiment of the present application. The embodiment of the present application is optimized based on the above embodiment.

[0035] As shown in FIG2 , the method of the embodiment of the present application includes the following steps:

[0036] S210, obtaining concentration data of dissolved gas in transformer oil.

[0037] S220 , performing normalization processing on the concentration data to obtain normalized concentration data.

[0038] For example, after obtaining the concentration data of the dissolved gas in the transformer oil, the concentration data is mapped to [0, 1] to obtain normalized concentration data.

[0039] This scheme is set up in this way so that various concentration data types can be normalized and then input into the optimized probabilistic neural network model, avoiding the situation where inconsistent concentration data leads to difficulties in model recognition.

[0040] S230, processing the normalized concentration data based on the optimized probabilistic neural network model to obtain the fault type of the transformer; the optimized probabilistic neural network model is determined according to the target smoothing factor parameter; the target smoothing factor parameter is determined from the candidate smoothing factor parameters through the artificial jellyfish search algorithm; the candidate smoothing factor parameter is the smoothing factor parameter that meets the fitness condition.

[0041] For example, the process of determining candidate smoothing factor parameters includes: calculating fitness values ​​corresponding to initial smoothing factor parameters; and determining a first preset number of initial smoothing factor parameters with the smallest fitness values ​​as the candidate smoothing factor parameters.

[0042] It should be noted that in the artificial jellyfish search algorithm, it is necessary to determine a population and then determine the optimal position of the jellyfish from the individuals in the population. In the embodiment of the present application, the candidate smoothing factor parameters corresponding to the above population are the individuals in the population, and the target smoothing factor parameters correspond to the optimal position of the jellyfish.

[0043] For example, an initial smoothing factor parameter is pre-generated, and based on a fitness value calculation function, a fitness value corresponding to the initial smoothing factor parameter is calculated. A first preset number of initial smoothing factor parameters having the smallest fitness value are determined as the candidate smoothing factor parameters. The first preset number can be determined based on actual conditions and is not limited in this embodiment of the present application.

[0044] Exemplarily, if the number of initial smoothing factor parameters is 2N, the first preset number may be N, that is, after determining the fitness values ​​corresponding to the initial smoothing factor parameters, the first N smoothing factor parameters with the smallest fitness values ​​are determined as candidate smoothing factor parameters.

[0045] For example, calculating the fitness value corresponding to the initial smoothing factor parameter includes: subtracting the label result of the training sample from the output result of the target probabilistic neural network model for the training sample to obtain a first parameter; the target probabilistic neural network model is a probabilistic neural network model determined based on the initial smoothing factor parameter; taking the square of the first parameter as the second parameter; and taking half of the second parameter as the fitness value corresponding to the initial smoothing factor parameter.

[0046] Exemplarily, the fitness value e(k) is calculated using the following formula:

[0047] Among them, y(k) is the output result of the target probabilistic neural network model for the target training sample, and y'(k) is the label result of the target training sample; the target probabilistic neural network model is a probabilistic neural network model determined based on the target smoothing factor parameter.

[0048] For example, the process of determining the initial smoothing factor parameters includes: generating a second preset number of original individuals based on chaotic mapping; generating reverse individuals corresponding to the original individuals based on reverse learning; and determining the original individuals and the reverse individuals as the initial smoothing factor parameters.

[0049] Among them, the second preset number can be determined according to actual conditions, and the embodiment of the present application does not limit this.

[0050] Exemplarily, a second preset number of original individuals are generated by Skew Tent chaotic mapping, and an inverse solution is found for each individual through an opposition-based learning (OBL) strategy to form an inverse individual corresponding to the original individual, and the original individual and the inverse individual are determined as initial smoothing factor parameters.

[0051] This solution uses chaotic mapping to generate a second preset number of original individuals, and generates reverse individuals corresponding to the original individuals based on reverse learning, which broadens the search area and increases the number of initial smoothing factor parameters, making the final target smoothing factor parameters more accurate.

[0052] For example, the process of determining the target smoothing factor parameters includes: updating the position of the passively moving jellyfish in a global search manner; updating the position of the actively moving jellyfish; determining the optimal position of the jellyfish based on the position of the passively moving jellyfish and the position of the actively moving jellyfish, and using the optimal position as the target smoothing factor parameter.

[0053] In the embodiments of the present application, passive motion refers to the type A motion of the jellyfish, that is, the motion of the jellyfish following the ocean current; active motion refers to the type B motion of the jellyfish, that is, the motion of the jellyfish around itself.

[0054] For example, when updating the position of a passively moving jellyfish, it is updated in a global search manner to avoid local optimal situations.

[0055] For example, the update formula for the passive motion jellyfish position is as follows:

[0056] Among them, X i (t+1) is the position of the jellyfish at the t+1th iteration, X i (t) is the position of the jellyfish at the tth iteration, γ is the motion coefficient, U b is the upper limit of the search space, Lb is the lower limit of the search space, r1 is rand(0,1), that is, a random number between [0,1], r2 is a random number between [0,1], g * is the current optimal smoothing factor parameter (i.e. the current target smoothing factor parameter), is the i-th candidate smoothing factor parameter at the t-th iteration, f i The jellyfish's luminous point.

[0057] The technical solution of the embodiment of the present application selects candidate smoothing factors that meet the fitness conditions from the initial smoothing factors, so that the candidate smoothing factors themselves are more suitable for the model than other smoothing factors, and the target smoothing factor parameters finally determined are better; and, in the process of determining the target smoothing factor parameters based on the artificial jellyfish search algorithm, the position of the passively moving jellyfish is updated in a global search manner, so that the update of the position of the passively moving jellyfish avoids the local optimal situation and achieves the global optimal situation, and the target smoothing factor parameters finally determined are the parameters at the optimal position of the jellyfish, so that the optimized probabilistic neural network model can more accurately identify the fault type.

[0058] For example, the parameters of the probabilistic neural network model are optimized by using the improved artificial jellyfish search algorithm to obtain the target smoothing factor parameters as follows:

[0059] In step A1, N individuals are generated through Skew Tent chaotic mapping to form the original population N1. Then, the reverse learning strategy is used to screen out N initial populations, and the reverse solution is found for each individual to form the reverse population N2. The original population N1 and the reverse population N2 are merged to obtain the initial smoothing factor parameters, and the fitness values ​​of the initial smoothing factor parameters are calculated. The top N with the smallest fitness values ​​are selected as candidate smoothing factor parameters.

[0060] Step A2: Evaluate the fitness value and determine the initial optimal position. The output mean square error value e(k) is used as the fitness function to evaluate the performance index of the initial probabilistic neural network model, that is,

[0061] Among them, e(k) is the fitness value, y(k) and y'(k) represent the actual output value and ideal output value of the initial probabilistic neural network model respectively.

[0062] Step A3, update the time control parameter c(t), that is

[0063] Where: t is the number of iterations, M a x iter is the maximum number of iterations.

[0064] Step A4, update the jellyfish position according to the ocean current, that is, X i (t+1)=X i (t)+rand(0,1)*(X * -β*rand(0,1)*u)

[0065] Where, X i (t+1) is the position of the jellyfish at the t+1th iteration, X i (t) is the jellyfish position at the tth iteration, X* is the optimal position of the current jellyfish population, u is the average position of all jellyfish, and β is the distribution coefficient, which is 3. It can be understood that when the number of iterations is 0, X* is the initial optimal jellyfish position determined in step A2.

[0066] Step A5: Update the movement type. The positions of Type A and Type B jellyfish are updated separately. The idea of ​​the global search phase of the Butterfly algorithm is introduced into the Artificial Jellyfish Search (AJS) algorithm to update the position of Type A jellyfish. The improved formula for updating the position of Type A jellyfish is as follows:

[0067] Where: U b and L b are the upper and lower limits of the search space, γ is the motion coefficient, which is 0.1, g* is the optimal solution so far, and f i is the luminous point of the jellyfish, r∈[0,1] is a random number.

[0068] The movement direction and updated position formula of Type B jellyfish are as follows:

[0069] Boundary Condition: When a jellyfish moves out of the bounded search area, it returns to the opposite boundary, as follows: i,d =(X i,d -U b,d )+L b (d), if X i,d >U b,d X' i,d =(X i,d -L b,d )+U b (d), if X i,d <L b,d

[0070] Where, X' i,d is the updated position of the jellyfish, X i,dis the position of the i-th jellyfish in d dimension, U b,d and L b,d are the upper and lower bounds of the search space.

[0071] Step A6, re-evaluate the fitness value and update the optimal position of the jellyfish.

[0072] Step A7, determine whether the current number of iterations meets the maximum number of iterations. If the current number of iterations meets the maximum number of iterations, proceed to step A8. If the current number of iterations does not meet the maximum number of iterations, return to step A3 and iterate again.

[0073] Step A8: Using the optimal position of the jellyfish as the target smoothing factor parameter of the probabilistic neural network model to obtain an optimized probabilistic neural network model.

[0074] Figure 3 is a schematic diagram of the structure of a transformer fault identification device provided in an embodiment of the present application. The device can execute the transformer fault identification method provided in any embodiment of the present application and has the corresponding functional modules and beneficial effects of the execution method. As shown in Figure 3, the device includes:

[0075] A concentration data acquisition module 310 is configured to acquire concentration data of dissolved gas in transformer oil;

[0076] The fault type determination module 320 is configured to process the concentration data based on an optimized probabilistic neural network model to obtain the fault type of the transformer; the optimized probabilistic neural network model is determined based on a target smoothing factor parameter; the target smoothing factor parameter is determined from candidate smoothing factor parameters through an artificial jellyfish search algorithm; the candidate smoothing factor parameter is a smoothing factor parameter that meets the fitness condition.

[0077] For example, the apparatus further includes a module for determining candidate smoothing factor parameters, and the module for determining candidate smoothing factor parameters includes:

[0078] A fitness value calculation unit is configured to calculate a fitness value corresponding to an initial smoothing factor parameter;

[0079] The candidate smoothing factor parameter determining unit is configured to determine a first preset number of initial smoothing factor parameters with the smallest fitness values ​​as the candidate smoothing factor parameters.

[0080] For example, the fitness value calculation unit is set to perform the following operations:

[0081] Subtracting the label result of the training sample from the output result of the target probabilistic neural network model for the training sample to obtain a first parameter; the target probabilistic neural network model is a probabilistic neural network model determined based on the initial smoothing factor parameter;

[0082] Taking the square of the first parameter as the second parameter;

[0083] One half of the second parameter is used as the fitness value corresponding to the initial smoothing factor parameter.

[0084] For example, the device further includes a module for determining an initial smoothing factor parameter, and the module for determining the initial smoothing factor parameter includes:

[0085] an original individual generating unit, configured to generate a second preset number of original individuals based on a chaotic map;

[0086] a reverse individual generating unit, configured to generate a reverse individual corresponding to the original individual based on reverse learning;

[0087] The initial smoothing factor parameter generating unit is configured to determine the second preset number of original individuals and the reverse individuals corresponding to the second preset number of original individuals as initial smoothing factor parameters.

[0088] For example, the device further includes a module for determining a target smoothing factor parameter, and the module for determining the target smoothing factor parameter includes:

[0089] A first position updating unit is configured to update the position of the passively moving jellyfish in a global search manner;

[0090] a second position updating unit, configured to update the position of the actively moving jellyfish;

[0091] The target smoothing factor parameter determination unit is configured to determine the optimal position of the jellyfish according to the updated passive motion jellyfish position and the updated active motion jellyfish position, and use the optimal position as the target smoothing factor parameter.

[0092] For example, the update formula for the passive motion jellyfish position is as follows:

[0093] Among them, X i (t+1) is the position of the jellyfish at the t+1th iteration, X i (t) is the position of the jellyfish at the tth iteration, γ is the motion coefficient, U b is the upper limit of the search space, L b is the lower limit of the search space, r1 is rand(0,1), that is, a random number between [0,1], r2 is a random number between [0,1], g * is the current optimal smoothing factor parameter, is the i-th candidate smoothing factor parameter at the t-th iteration, f i The jellyfish's luminous point.

[0094] For example, the device further includes a normalization processing unit configured to perform normalization processing on the concentration data to obtain normalized concentration data;

[0095] The fault type determination module 320 is configured to process the normalized concentration data based on the optimized probabilistic neural network model.

[0096] A transformer fault identification device provided in an embodiment of the present application can execute a transformer fault identification method provided in any embodiment of the present application, and has functional modules and beneficial effects corresponding to the execution method.

[0097] FIG4 shows a block diagram of an electronic device 10 that can be used to implement an embodiment of the present application. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0098] As shown in FIG4 , the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 and a random access memory (RAM) 13, that is communicatively connected to the at least one processor 11. The memory stores a computer program that can be executed by the at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are connected to each other via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0099] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0100] The processor 11 can be a variety of general-purpose and / or specialized processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any other suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the transformer fault identification method.

[0101] In some embodiments, the transformer fault identification method can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the transformer fault identification method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to execute the transformer fault identification method in any other appropriate manner (e.g., via firmware).

[0102] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0103] Computer programs for implementing the methods of the present application may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0104] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. A computer-readable storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. A computer-readable storage medium can be a non-transient computer-readable storage medium.

[0105] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a cathode ray tube (CRT) or a liquid crystal display (LCD) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0106] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0107] A computing system may include a client and a server. The client and server are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host, a host product within a cloud computing service ecosystem that avoids the management difficulties and limited scalability inherent in physical hosting and VPS services in related technologies.

[0108] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the multiple steps described in this application can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of this application can be achieved. This is not limited herein.

Claims

1. A transformer fault identification method, comprising: Obtain the concentration data of dissolved gas in transformer oil; The concentration data is processed based on the optimized probabilistic neural network model to obtain the fault type of the transformer; wherein the optimized probabilistic neural network model is determined according to a target smoothing factor parameter; the target smoothing factor parameter is determined from candidate smoothing factor parameters by an artificial jellyfish search algorithm; the candidate smoothing factor parameters are smoothing factor parameters that meet fitness conditions.

2. The method according to claim 1, wherein: The process of determining the candidate smoothing factor parameters includes: Calculate the fitness value corresponding to the initial smoothing factor parameter; A first preset number of initial smoothing factor parameters with the smallest fitness values ​​are determined as the candidate smoothing factor parameters.

3. The method according to claim 2, for example, the step of calculating the fitness value corresponding to the initial smoothing factor parameter comprises: Subtract the label result of the training sample from the output result of the target probabilistic neural network model for the training sample to obtain a first parameter; The target probabilistic neural network model is a probabilistic neural network model determined based on the initial smoothing factor parameters; Taking the square of the first parameter as the second parameter; One half of the second parameter is used as the fitness value corresponding to the initial smoothing factor parameter.

4. The method according to claim 2, for example, the process of determining the initial smoothing factor parameter comprises: Generating a second preset number of original individuals based on the chaotic map; Generate a reverse individual corresponding to each original individual based on reverse learning; The second preset number of original individuals and the reverse individuals corresponding to the second preset number of original individuals are determined as initial smoothing factor parameters.

5. The method according to claim 1, for example, the process of determining the target smoothing factor parameter comprises: Update the position of passively moving jellyfish by global search; Update the position of actively moving jellyfish; The optimal position of the jellyfish is determined according to the updated passive motion jellyfish position and the updated active motion jellyfish position, and the optimal position is used as a target smoothing factor parameter.

6. The method according to claim 5, wherein: The updating formula of the passive motion jellyfish position is as follows: Among them, X i (t+1) is the position of the jellyfish at the t+1th iteration, X i (t) is the position of the jellyfish at the tth iteration, γ is the motion coefficient, U b is the upper limit of the search space, L b is the lower limit of the search space, r1 is rand(0,1), that is, a random number between [0,1]. r2 is a random number between [0,1], g * is the current optimal smoothing factor parameter, is the i-th candidate smoothing factor parameter at the t-th iteration, f i The jellyfish glows.

7. The method according to claim 1, for example, after obtaining the concentration data of dissolved gas in transformer oil, further comprises: Normalizing the concentration data to obtain normalized concentration data; The processing of the concentration data based on the optimized probabilistic neural network model includes: The normalized concentration data is processed based on the optimized probabilistic neural network model.

8. A transformer fault identification device, comprising: A concentration data acquisition module, configured to acquire concentration data of dissolved gas in transformer oil; A fault type determination module is configured to process the concentration data based on an optimized probabilistic neural network model to obtain the fault type of the transformer; wherein the optimized probabilistic neural network model is determined according to a target smoothing factor parameter; the target smoothing factor parameter is determined from candidate smoothing factor parameters by an artificial jellyfish search algorithm; and the candidate smoothing factor parameters are smoothing factor parameters that meet fitness conditions.

9. An electronic device, comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the transformer fault identification method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a processor to implement the transformer fault identification method according to any one of claims 1 to 7 when executed.

Images